(Original post: August 8, 2015)
As I have been facing Kelsey’s and my recent training challenge (Gamblers), I’ve reflected a lot on all that I’ve heard about training distances and gamblers. I’ve heard things like:
“You don’t need to train gambles, train value of equipment.” “You just need good solid verbals.” “You need to train gambles.” “There are simply some dogs that are good at gamblers and others are good at snooker. You can’t have both.”
It really got me to thinking. I think there are bits of truth to all of the above, but perhaps there’s a better way of looking at it. In this Training Thought, I link together all that I know about training happy, confident dogs on equipment and principles of learning and driving accurate lines. In summary, it comes down to a combination of the dog’s: value for equipment, understanding of directionals and value for completing those directionals, understanding of the behaviors / paths in different habitats or contexts (DASH principle from Say Yes), and arousal level. The composition of the mix may differ depending on the dog and as such a weakness in one area makes the others more relevant. Let’s start off with value first.
1) Value for Equipment
Value for equipment refers to whether or not the dog finds performing the equipment to be intrinsically rewarding. This is relevant as the value of performing the equipment needs to be high and of at least equal value to being beside the handler. In gamblers there is a conflict between the value of being beside Mom/Dad, versus the value of playing with the equipment. For some dogs, the joy of running and doing equipment is extremely rewarding. For others, it is not as rewarding as being right beside the “cookie dispenser.” The former dogs might end up having more challenges in other types of games, like snooker, where the value beside the handler needs to be higher than the value of grabbing equipment.
The latter dogs learn that the further they are from Mom/Dad, the less frequently they will get their A+ rewards. So if the value of doing the equipment cannot at least balance the value of being beside Mom/Dad, then it will be very challenging to get any distance. So what’s the fix for this? Ironically, it isn’t necessarily upping the distance, but rather improving reward delivery mechanics and upping the reinforcement rate of performing equipment. Increase and improve the transfer of value to the equipment. This means, close, frequent payouts. However, the payout must NOT be on/from/delivered by hand to the dog from the Handler as that increases the value of the Handler’s proximity.
Bob Bailey spoke of training dolphins for war training. They worked intensely in closed spaces with the dolphins to up the reinforcement rates and the number of trials (hence learning) before ever attempting any distances. This up-close training was far more effective. But the key here is mechanics!!! Rewards must magically appear on line immediately after dog completes the equipment and before they check back into Mom/Dad! For example, Susan Garrett has a fantastic one-jump exercise to up the value of the jump on her DVD Success with One Jump.
2) Understanding of Performance – Independent of Handler
The second step to easier gamblers is through answering the question, “Does the dog understand how to perform the equipment, independent of the handler?” The second part is really the critical part. So often, handlers infuse themselves into the performance picture for the dog. For example, the dog may think “I know to take a jump when my human has their arm up and flicks upwards when they are right beside that jump, and they say, ‘Jump,’ right at that obstacle.” If anything changes in that picture, the dog appears to not know how to jump. That simply means that the handler has been babysitting and over-prompting the behavior.
A fantastic way to fix this is to alter the handler’s body mechanics. Tie down the arms if they can’t stop waving the arm, have the handler race the dog, have the handler sit on the ground and ask for the behavior, have the handler twirl and skip. By making the handler’s body unpredictable, it allows the dog to understand better the true meaning of “Jump.”
Again, Susan Garrett has a fantastic game called “Can You In A Box?” (inspired from the Dr. Seuss’ story Green Eggs and Ham). She did an experiment with a group of Say Yes Campers, where one team tried to train the farthest distances, and the other group try to train the most whacky-weird-crazy thing they could possibly do (handstands, somersaults, you name it). The next day, they tested to see which group could get the farthest distance on the jump. The Can You In A Box group won, hands-down. (Pardon the pun.)
I have even found with Kelsey that my “plank while planking” challenge with a friend in the UK (I do a plank (think static push-up) while Kelsey performed the dog walk with a running contact helped. It took her a bit of time to realize criteria were still being expected, but after that, I was suddenly able to get a 30’ lateral DW with a happy dog on a single attempt! That truly shocked me, and she got an A+ reward for that awesomeness!
3) Value and Understanding of Verbals
So many dogs rely on the handler’s position and motion to inform them where they are heading next; however, in gamblers, the handler is unable to be in a close position, and so many handlers end up frozen (not moving except flailing arms) due to the gamble line. At one trial in Kingston, I watched a rough coat collie perform a Masters Gamblers with verbals. Even though the dog was a little confused on where she was supposed to be going, the handler was able to direct her, like a remote controlled car, through the gamble. It really was brilliant to watch! After the run, the handler and I laughed and marvelled at her incredible understanding. Even with Kelsey I have watched her confusion and frustration level diminish significantly as her understanding of directional verbals has improved. I have seen many gamble set-ups that would be a lot easier to achieve in a fluid way if the dog has a solid understanding of wraps and left/right verbals. However, this factor alone is insufficient to overcome a lacking in the other factors.
4) Understanding of Performance – Context Independent
This one is really the bit that got me thinking about the debate on whether or not you need to “train gambles, per se.” When we train any new behavior, we are very aware that the dog’s understanding of that behavior is not truly understood until it has been trained in multiple habitats (a.k.a., environments and contexts). We know this is true because of the popular conundrum of “But my dog can do this at home!” A dog doesn’t really have a solid Down until they can do it in the kitchen, tv room, backyard, front yard, out on a walk, in the forest, in a field, at a school yard, at a shopping mall, at the training arena alone, at the training arena with dogs present, at the training arena with other dogs present who are running around and having a blast. All of these are different contexts.
In terms of a line of equipment, we know it needs to be trained in the arena, in the backyard, at the park, at a trial site, at another trial site, with different looking equipment, with different numbers of bars, with and without number cones. All of these nuances add to the picture and habitat. I have been looking at gamblers and there are different types of gambles, some of which alter the picture of the line, for example, layering. By putting a dog walk, or tunnel, or even a jump parallel to the dog’s path, and asking the dog to complete that line (without taking the layered obstacle), alters the habitat. For some dogs it is like leaving them in your backyard, going to your neighbour’s yard, and asking for a behavior through the fence. Easy? Not easy? The answer is habitat.
Another typical gamble set-up is a box, where the dog must drive between two wrong-course obstacles (like a jump and teeter) to take the correct one at a distance (like a tunnel). Again, this is a different context of that original line of simply driving to the tunnel. Until the dog has been exposed to these different habitats, some dogs may be more sensitive to the Habitat and lose their confidence or appear to not know the behavior.
To that end, it isn’t “training gambles” per se, but rather, improving the dog’s understanding of verbals and their path in a different context. I had the hardest time figuring this part out, because I truly didn’t believe that one should train gambles. In all my years, I have never specifically trained gamblers. My past two agility partners seemed to have no particular issue with gambles, but Kelsey has had issues despite the fact that Kelsey has high value for equipment, she can do the equipment independent of me, and she has solid (80+% accuracy) on her verbals. When we would be faced with a challenging new habitat or presentation of the path, she was notably unsure and would question the commands and eventually just grab the closest obstacle (almost as is if to end the uncertainty). However, after I revisited the first three factors, I started treating the different types of gambles as just another new H (from DASH – Drive, Accuracy, Speed, Habitat; Susan Garrett). I would shape and reward her choices, and we are now back on the road to having a confident dog in gambles. We still need more work on some of the H’s but we are a lot further ahead than before.
This brings me to the final component. I really started to wonder why some dogs (like Kelsey) were so sensitive to the new H’s, while others seemed oblivious to them, and would charge on beautifully. I believe the answer is in the fifth aspect, arousal level.
5) Arousal Level
We know that there is an ideal arousal level for peak performance according to the Yerkes-Dodson law (Yerkes & Dodson, 1908). When a dog is under-aroused then the dog can become attentive to trivial stimuli and notice more in their environment. When a dog is over-aroused, then the dog doesn’t take in sufficient stimuli (they become myopic in their focus) to perform well. For example, with Kelsey, I have noticed that scents and birds can either be a non-issue for her (they can fly overhead in an indoor facility without altering her performance), or the mere scent of them can take her off the path for a good sniff (same day, same ring, same birds). These two behaviors are entirely related to her arousal level. If it is too low, I can guarantee you a sniff shall happen! If it is too high, then I have very little control on the field and contacts / tables / start lines / turn cues are hard to achieve. But if it is just right, then it is magic with just the two of us doing our dance together. I have also noticed with Kelsey and my students’ dogs, that arousal can shift over the course of a run. For some, they may run high and get more and more aroused over the course of the run, especially if there are no control points on the course. For those dogs, straight-line-grab-what-you-see gambles would be very easy for them, but gambles which require them to also listen to handler verbals would become more challenging due to over-arousal.
For others, they may go lower and lower in arousal, maybe due to fatigue, maybe due to decreased confidence and occurrence of questions while on course. For these dogs, by the time they get to the main gamble, they are into under-arousal and habitat becomes HABITAT!!!
So to address the arousal issue, I suggest that for those who have dogs who run hot and become overly aroused, adding some control into the opening might actually help moderate the arousal level for the dog. The amount and type of control will depend on the dog’s arousal and confidence on those control factors (like a wing wrap, or contacts). For those who have dogs who become under-aroused, then flow and confidence boosting openings become even more critical. For Kelsey, until I know her confidence on different gamble contexts is high, we tend to be more successful with fun, fast, flowing openings to minimize the probability that the change of habitat is even noticed.
So there you have it. Gamblers definitely is a multi-tiered issue for dogs and handlers, but really the neat thing about them is that they test what you have been training with your dog, and test your dog’s true understanding of driving the line you have requested. I should note that at no point in here do I discuss the handler’s role in handling the gamblers. I have always told my students to ignore the gamble line, and find their own handling line (away from the gamble line) to give them buffer, so they aren’t caught, stuck, flailing their arms at their dog. However that’s another Training Thought topic. This one is on the dog training and the dog’s understanding components of gamblers. 🙂
I hope this helps! Happy training!